Data Analytics tools provide Data Scientists and Data Engineers with the instruments to find patterns in data and provide business insights to Executives and other stakeholders. Data Analytics Notebooks are programmes that fast track this process as Data Scientists can very quickly develop applications to analyse the data, visualise the results and roll the successful patterns out into production. This shortens the pattern analysis cycle and reduces the hit and miss common in the development of analytics models.
The Jupyter Notebook is a well-known software application that has been around for a while. It has been used by some big companies like Google and NASA. It’s best suited for working with data that can fit into memory. It was developed in 2012, an evolution of the iPython Notebook, which only supported Python language as a notebook engine.
Jupyter is more flexible and supports a number of programming languages such as Python, Scala and R. Since it’s open-source, it has a big community and a lot of additional software and integrations.
On the other hand, Apache Zeppelin was developed in 2013 by Apache Foundation as part of the Hadoop landscape. While it is also open-source, its community is only a small fraction of Jupyter’s. It’s more suited for data that is distributed across a Hadoop cluster. One of Zeppelins strengths is the creation of dashboards and multi-user sharing.
Another advantage of Zeppelin is that it is part of the Hadoop landscape and integrates well with other Hadoop applications such as Spark, Pig, Hive and others.
Appearance
Both UIs are similar, but Zeppelin’s biggest advantage is that it allows the combination of multiple paragraphs into one line. It also has a built-in simple data visualisation tool for some interpreters and a table output that allows sorting out-of-the-box.
Jupyter’s code editor and paragraph editor seem to be much more effective though, with more hot keys and a great auto-completion feature. Another plus of this software lies in the large number of Python libraries for visualizing data that support output of pictures and other interactive content directly in paragraph outputs.
Multi-User Support & Extensions
Jupyter doesn’t support multi-user configuration by default, but by installing Jupyterhub, an additional service can be added that accepts client connections, authenticates them, and starts a separate Jupyter server. This is not a great solution if a large number of users need to be accommodated because it may result in loading overheads on the physical server.
Zeppelin supports multi-user configuration via LDAP/Active Directory connectivity and specifically defined security groups. It uses only one server process, authenticating users in the configured system before allowing further access. Zeppelin provides a number of interpreters that provide library functions. The most common are Spark and Scala that open the tool to a large number of specialised match and statistics libraries. The same is true for Jupyter that has more extensions available to use because of its huge community and its stand alone nature.
In conclusion, Zeppelin is the better tool to use if the data scientist develops in the Hadoop world. It provides good integration with other Hadoop systems such as Spark, Pig and others and streamlines the development for Spark applications. It provides a better integration of larger teams, but it seems more geared towards enterprise users, having great LDAP integration, permissions management, and so on.
Jupyter requires less overhead in the setup and productionisation of developed patterns due to the stand alone nature. Due to the large number of extensions and integrations, specifically into Machine Learning and AI frameworks it has developed into the more popular choice among analytics users.
Discuss Jupyter and Zeppelin comparison or the pros and cons of each platform further with Fusion Professionals.
